The proportion of smokers who are heavy smokers (25 or more cigarettes per day) has increased. Research to develop methods to promote cessation among heavy smokers is needed since they are at increased risk for disease and death. Several factors influence the design of interventions for heavy smokers. Heavy smoking, compared to light smoking, probably reflects physical dependence processes more directly. Treatments for heavy smokers may need to manage physical dependence in order to provide protection from the effects of withdrawal. Survey data indicate that most smokers wishing to quit desire self-help, minimal contact cessation programs. The integration of pharmacologic and psychological strategies within a self-help treatment program for heavy smokers offers a new challenge for smoking researchers. The proposed project is a randomized experiment in a defined population comparing the effectiveness of a monetary incentive alone or with a self-help psychological treatment, a pharmacologic treatment, or both treatments combined in producing sustained abstinence for 6-12 months. Virtually all smoking cessation studies have been conducted with volunteers who may not accurately represent the general population of smokers. This study will identify heavy smokers in a representative sample of a defined population and explore their demographics, knowledge, attitudes, smoking and health history, and an estimated 2,800 smokers who will be invited to participate in the trial. All will be offered a $100 cash payment if they meet the following conditions: a) quit smoking for 24 hours with expired-air CO verification; b) agree to participate in a self-help relapse prevention program incorporating nicotine gum and coping skills training materials; c) demonstrate abstinence via a negative saliva cotinine value obtained at 3 and 6 month evaluations. Telephone prompts and mail-ins will be used to boost adherence to the relapse prevention protocol. Smokers who do not wish to participate or who fail to quit for 24 hours will be assigned to a Monitoring condition. Half of these subjects will be re-invited to participate at two month intervals. The other half will serve as a control. We will assess reported smoking, perceived stress and social support. Outcome measures will be collected at 0,3,6,12 and 24 months on all participants, including the Monitoring Group, and self-reported smoking status will be confirmed by expired-air CO and saliva cotinine. Power calculations show that we can detect 12% differences in abstinence rates.